Storage and Server Provisioning for Virtualized and Geographically Dispersed Data Centers

1. A method of migrating an application, comprising: installing the application on a first server by setting up a first virtual machine running a guest operating system at a first data center, the application accessing a first logical unit (LU) on a first storage system at the first data center, wherein data used by the application is stored in the first LU; maintaining a global resource manager at the first data center, the global resource manager comprising a list of one or more resources available to a plurality of remote data centers and status information for each of the one or more resources, the list of one or more resources including information on one or more logical units (LUs) available to the plurality of remote data centers, wherein the plurality of remote data centers each include a storage system having a plurality of storage devices for providing the one or more LUs; initiating migration of the application by searching for a candidate remote data center from among the plurality of remote data centers; consulting the global resource manager to determine whether a replica of the first LU already exists in any of the plurality of remote data centers, wherein the replica of the LU is a LU having a copy of the data used by the application; when a replica of the first LU exists, determining whether the remote data center at which the replica exists has server resources that meet requirements for migration of the application, and when the remote data center at which the replica exists has server resources that meet the requirements for migration of the application, identifying the remote data center at which the replica exists as the candidate remote data center for receiving the migration; when a replica of the first LU does not exist, determining whether any of the remote data centers have storage resources and server resources that meet requirements for migration of the application and the data used by the application, and when one of the remote data centers has storage resources and server resources that meet the requirements for migration of the application and the data used by the application, identifying the remote one of the remote data centers as the candidate remote data center for receiving the migration; and in response to identifying a candidate remote data center, migrating the application and the guest operating system to a second virtual machine installed on a second server located at the candidate remote data center.

2. The method according to claim 1 , further including the step of initiating the migration by first searching said local data center for available resources by determining if a replica of said first LU exists at the local data center and is accessible by available server resources that meet requirements for migration of the application; when a replica of the first LU does not exist at the local data center, determining whether the local data center has storage resources and server resources that meet requirements for migration of the application and the data used by the application, and when a replica of the first LU does not exist at the local data center and the local data center does not have storage resources and server resources that meet requirements for migration of the application and the data used by the application, searching for available resources in the plurality of remote data centers.

3. The method according to claim 1 , further including the step of creating said first LU as a virtual LU, wherein data stored to said virtual LU on said first storage system is stored by said first storage system to an actual LU in another storage system that maps to the virtual LU, said other storage system being located at the local data center and storing the data to physical storage devices from which the actual LU is configured.

4. The method according to claim 1 , wherein said step of identifying a candidate remote data center includes examining power consumption information for the candidate remote data center to determine whether power consumption for the candidate remote data center is within a specified limit.

5. The method according to claim 1 , further including the step of replicating the first LU to the candidate remote data center when a replica of the first LU is not already present at the candidate remote data center using a remote copy function, so that the replica of the first LU at the candidate remote data center is synchronized with the first LU when the application is suspended at the local data center and the migrated application at the remote data center is able to use the replica of the first LU to take over for the suspended application.

6. An information system comprising: a plurality of data centers located geographically remote from each other, the data centers each including one or more servers and one or more storage systems, the data centers being in communication with each other via a network, wherein the plurality of data centers each include a storage system providing logical units (LUs) which hare configured by a plurality of storage devices; a first virtual machine running on a first server at a first data center, the first virtual machine having a guest operating system with an application installed thereon, the application accessing a first logical unit (LU) on a first storage system at the first data center, wherein data used by the application is stored in the first LU; and a global resource manager at the first data center that collects lists of resources, including LUs, and statuses of resources, from each of the plurality of data centers to create and maintain a list of all the resources of all the data centers, wherein, in response to initiating a migration of the application, the first data center is configured to consult the global resource manager to determine whether a replica of the first LU already exists in any of the plurality of remote data centers, wherein the replica of the LU is a LU having a copy of the data used by the application, wherein when a replica of the first LU exists, the first data center determines whether the data center at which the replica exists has server resources that meet requirements for migration of the application, wherein, when the data center at which the replica exists has server resources that meet the requirements for migration of the application, the first data center identifies the data center at which the replica exists a candidate remote data center for receiving the migration of the application, wherein, when a replica of the first LU does not exist, the first data center determines whether any of the data centers have storage resources and server resources that meet requirements for migration of the application and the data used by the application, wherein, when one of the remote data centers has storage resources and server resources that meet the requirements for migration of the application and the data used by the application, the first data center identifies the one of the data centers for receiving the migration, and wherein, in response to identifying a candidate remote data center, the first data center is further configured to migrate the application and the guest operating system to a second virtual machine installed on a second server located at the candidate remote data center.

7. The information system according to claim 6 , wherein said first data center is configured to initiate the migration of the application when a load on the first data center exceeds a predetermined load threshold set for the first data center.

8. The information system according to claim 6 , wherein said first data center is configured to initiate the migration of the application when an amount of power consumed at the first data center exceeds a predetermined power consumption threshold set for the first data center.

9. The information system according to claim 6 , wherein said first LU is a virtual LU configured on said first storage system, said virtual LU mapping to an actual LU on a second storage system at said data center, wherein data stored to said virtual LU on said first storage system is stored by said first storage system to the actual LU in the second storage system in the first data center, which stores the data to physical storage devices.

10. The information system according to claim 6 , wherein said first data center is configured to identify a candidate data center for receiving migration of the application, at least in part, by examining power consumption information for the data centers to determine whether the power consumption for the candidate data center is below a specified threshold.

11. The information system according to claim 6 , wherein said first data center is configured to replicate the first LU to the destination data center when a replica of the first LU is not already present at the destination data center using a remote copy function, so that the replica of the first LU at the destination data center is synchronized with the first LU, such that, when the application is suspended at the local data center, the migrated application at the destination data center takes over for the suspended application.

12. An information system comprising: a local data center having one or more local servers in communication with one or more local storage systems; a plurality of remote data centers located geographically remotely from each other and the local data center, the remote data centers each including one or more remote servers and one or more remote storage systems in communication with each other, each remote storage system providing logical units (LUs) which are configured by a plurality of storage devices; the local data center and the plurality of remote data centers being in communication with each other via a network; a first virtual machine running on a first server at the local data center, the first virtual machine having a guest operating system with an application operational thereon, the application conducting input/output (I/O) operations to a first logical unit (LU) on a local storage system at the first data center, wherein data used by the application is stored in the first LU; and a global resource manager at the local data center that collects lists of resources, including LUs, and statuses of resources, from each of the plurality of remote data centers to create and maintain a list of all the resources of all the data centers, wherein, when migration of the application is initiated, the local data center is configured to determine whether the local data center has server resources that meet requirements of the migration of the application; wherein, when the local data center does not have server resources to meet requirements of the migration of the application, the local data center is configured to consult the global resource manager to determine whether there is a replica of the first LU already existing in any of said plurality of remote data centers, wherein the replica of the first LU is a LU having a copy of the data used by the application, wherein, when a replica of the first LU exists, the local data center determines whether the data center at which the replica exists has server resources that meet requirements for migration of the application, wherein, when the data center at which the replica exists has server resources that meet the requirements for migration of the application, the local data center identifies the data center at which the replica exists as a candidate data center for receiving the migration, wherein, when a replica of the first LU does not exists, the local data center is configured to determine whether any of the plurality of remote data centers have storage resources and server resources that meet requirements for migration of the application and the data used by the application, wherein, when one of the data centers has available storage resources and server resources that meet the requirements for migration of the application and the data used by the application, the local data center is configured to identify one of the data centers as a candidate data center for receiving the migration, and wherein, in response to identifying a candidate remote data center, the first data center is further configured to migrate the application and the guest operating system to a second virtual machine installed on a second server located at the candidate remote data center.

13. The information system according to claim 12 , wherein said local data center is configured to initiate the migration of the application when an amount of power consumed at the local data center exceeds a predetermined power consumption threshold or when a load on the local data center exceeds a predetermined load threshold.

14. The information system according to claim 12 , wherein said first LU is a virtual LU configured on said first local storage system said virtual LU mapping to an actual LU on a second local storage system at said local data center, wherein data stored to said virtual LU on said first local storage system is transferred by said first storage system to the actual LU in the second local storage system in the local data center, which stores the data to physical storage devices.

15. The information system according to claim 12 , wherein said local data center is configured to identify a candidate data center for receiving migration of the application by examining power consumption information for the data centers to determine whether the power consumption for the candidate data center is below a specified threshold.

16. The information system according to claim 12 , wherein said local data center is configured to replicate the first LU to the destination data center when a replica of the first LU is not already present at the destination data center using a remote copy function, so that the replica of the first LU at the destination data center is synchronized with the first LU, such that, when the application is suspended at the local data center, the migrated application at the destination data center uses the replica of the first LU to take over for the suspended application.